QA GoGo Open the app

Requirements in.
Test cases out.

QA GoGo reviews your BRD, writes the test plan and test cases, checks what is still uncovered, and drafts the automation framework. You review every step before it moves on.

Open the app Explore the sample project

Free to open. There is no account: you add your own Claude API key and pay Anthropic only for what you generate.

BR-3

A promo discount cannot exceed 50% of the subtotal.

TC-005Discount is capped at 50% of the subtotalFunctionalHigh
TC-006Expired promo code is rejectedNegativeMedium
TC-014Cap still holds when the code is worth more than the cartBoundaryHigh
Question for the business
Is the 50% cap applied before or after the delivery fee?
Example output from the sample project, not a live result.

Built for people who write the tests

Use it where the blank page costs you the most time.

QA engineers

Turn a new BRD into a first-draft plan and case list in one sitting, then spend your time on judgement instead of typing.

QA leads

Standardise the format. Edit the prompts once so every plan and case follows your template and naming rules.

Freelancers and consultants

Start client work faster. Nothing is stored on a server, and you control which model and key are used.

See the app at work

Screens from the built-in sample project, a food-ordering app. The content is demo data, not live AI output.

See a full sample QA pack

Find the gaps before you test

The review lists unclear, missing and conflicting requirements, each with a question to take back to the business.

  • Severity on every finding
  • Questions you can paste into a ticket
qagogo / app
The requirement review step showing findings with severity

Test cases you can filter and export

Every case links to a requirement. Filter by priority, add more with one click, then export to CSV or Jira.

  • Priority and type counts at a glance
  • Refine adds cases, it never replaces yours
qagogo / app
The test cases step with counts, filters and a table of cases

Check coverage against the BRD

See which requirements are covered, partly covered or missing, and draft cases for the gaps.

  • Works on cases you wrote yourself
  • One click to draft the missing ones
qagogo / app
The coverage check step

See what is not tested yet

Every test case is linked to a requirement. The coverage check shows the requirements that have no test, so gaps surface before launch instead of after.

Illustration based on the sample project.

BR-1Login lockout2 test casesBR-2Delivery fee2 test casesBR-3Promo cap2 test casesBR-4Cash limitNo test case yetTC-001TC-002TC-003TC-004TC-005TC-006

What each step gives you

Run the whole workflow in order, or use one tool on its own.

ToolWhat you getExport as
Requirement review

Ambiguities, gaps and contradictions in your BRD, each with the question to ask the business.

On screen, Markdown
Test plan

Scope, approach, a traceability matrix, risks and exit criteria, with open questions called out.

Markdown
Test cases

Steps, test data and expected results, each linked to a requirement. Filter by priority and add more with one click.

CSV, Jira import, Markdown
Coverage check

Which requirements are covered, partly covered or missing, plus draft cases for the gaps.

Markdown
Automation framework

Folder structure, page objects and starter scripts for Playwright, Cypress, Selenium or API tests.

Markdown
Bug reports

Rough notes turned into a complete report with steps, severity and what is still missing.

Markdown

The same work, less typing

What changes when the first draft is generated and you do the reviewing.

By handWith QA GoGo
Requirement reviewRead the BRD, note questions as you go, easy to miss a conflict.A list of ambiguities and gaps, each with the question to ask.
Test planStart from an old template and rewrite it.A plan with a traceability matrix, risks and exit criteria.
Test casesWrite each case, steps and expected results included.Cases linked to requirements, ready to filter and export.
CoverageCross-check cases against the BRD in a spreadsheet.Covered, partial and missing, with drafts for the gaps.

Every output is a draft. You still review, correct and own the result.

How it works

Each step feeds the next. Stop, edit or skip ahead whenever you like.

  1. Your BRDPaste requirements or user stories
  2. 1ReviewSpot gaps and questions before anyone writes a test
  3. 2Test planScope, risks and a traceability matrix
  4. 3Test casesSteps, data and expected results per requirement
  5. 4AutomationFramework structure and first scripts

You review and edit at every step before the next one starts.

Use any time
Coverage checkcases vs BRD Bug reportnotes to report ExportMarkdown, CSV, Jira CSV Refineadd cases, shorten, focus

You stay in control of the output

The AI writes the first draft. Your team decides what ships.

Edit the instructions. Every step shows its system prompt. Change the role, the format or the rules, or add your own, such as Gherkin or your naming standard.
Refine without starting over. Ask for negative cases, a shorter plan or API tests. Test cases are added to your table instead of replacing it.
It flags what it does not know. Prompts tell the model not to invent requirements and to list unclear points as questions.
Review is part of the flow. Results are drafts. Each step ends with a prompt to check the output before you continue.

Your key, your documents, no middleman

QA GoGo has no server. Everything runs in your browser.

Your browser BRD, key, results Claude API api.anthropic.com direct request
You use your own Claude API key. Paste it once in Settings. Anthropic bills usage to that key, and the site itself costs nothing.
Your documents are not stored. Results stay in the open tab and are cleared when you close it. Download what you want to keep.
You control the spend. Pick the lower-cost Haiku model, and set a monthly limit in the Claude Console before you start.
Check your policy first. Anthropic's API terms apply to what you send. If your BRDs are confidential, confirm they can be sent to Claude.

Questions

Do I need an account?

No. Open the app, add your Claude API key in Settings, and start. There is nothing to sign up for.

What does it cost?

The site is free. Each result is generated by Claude through your API key, and Anthropic charges that key for the usage. A full run writes several long documents, so cost depends on the size of your BRD and the model you choose.

Where does my BRD go?

From your browser straight to the Claude API. QA GoGo never receives your documents or your key. Your prompt edits, and your key if you choose to remember it, are saved in this browser only.

How accurate is the output?

Treat it as a strong first draft. The prompts tell the model to stay within your BRD and to flag gaps, but it can still be wrong or miss a rule. Review each step before you rely on it.

Which models can I use?

Claude Sonnet 5.5 is the default. In Settings you can switch to Haiku 5.5 for lower cost or Opus 5.5 for the highest quality.

What can I export?

Markdown for every step, a CSV of test cases, and a CSV in the layout Jira's CSV import expects. You can also copy any result.

See it on a sample project

The sample loads a food-ordering BRD with a finished review, plan, test cases and coverage check. No key needed to look around.

Explore the sample project Start with your own BRD
Product manager edition

A version for writing and sharpening requirements is planned. Tell us if you want it.

Notify me